Bayes’ Theorem at work at the Event Horizon Telescope

Katie Bouman
Katie Bouman

This beautiful and smiling girl is my “hero of the year”. Her name is Katie Bouman and she is an Electrical Engineer and a Ph.D. Candidate at MIT.

http://people.csail.mit.edu/klbouman/

https://www.ted.com/speakers/katie_bouman

Since 2014 she is working at the image analysis system used by the Event Horizon Telescope:

http://www.eventhorizontelescope.org/

https://www.wikiwand.com/en/Event_Horizon_Telescope

Katie became my hero of the year in November 2106, when I watched this TED Talk:

 

The reason why I was excited by her talk is not related to the possibility to¬†take a shot of a black hole. It is the fact that she is using a special “implementation” of Bayes’ Theorem to do so.

Bayes’ Theorem is one of my preferred “tool” and one of my long-standing interests. Bayes Theorem is a well-known and largely appreciated tool used by computer scientist to perform a large set of different task, from email spam fighting to computer vision. I met it years ago, when first Bayesian spam filters landed on the market, but I really understood it, and I really begun to appreciate it, only after having read this book:

Now, I consider Bayes’ Theorem as one of the most important and most useful tool for the development of Artificial¬†Intelligence software and – much more important – as the main and most important “algorithm” used by human and animal intelligence. Yes, you heard it: Bayes’s Theorem is arguably the main “algorithm” used by your brain to perform its reasoning, at any level (movement control, situation analysis and so on).

I will not bore you with all of the details. Just get a free copy of this fantastic technical manual and read it:

The free, electronic version is here:

Think Bayes

Of course, you can also get a shortcut and read the page of Wikipedia devoted to Bayes’ Theorem:

https://www.wikiwand.com/en/Bayes’_theorem

Katie Bouman is using Bayes’ Theorem in a very sophisticated way and it can be hard to recognize the old Reverend Bayes under all of this modern mathematics but he is still there.

In my humble opinion, this way of doing VLBI and image reconstruction is bound to a great success. I expect to see it used in many fields in the near future, from astrophotography to forensic analysis.

Stay tuned…

Alessandro Bottoni

Bayes’ Theorem at work at the Event Horizon Telescope